矩阵损失下带约束的回归系数Minimax可容许性的特征

THE MINIMAX ADMISSIBILITY CHARACTERIZATION OF HOMOGENEOUS LINEAR ESTIMATES WITH RESPECT TO RESTRICTED REGRESSION COEFFICIENT UNDER MATRIX LOSS FUNCTION

在矩阵损失下给出了带约束回归系数线性估计在齐次线性估计类中是Ｍｉｎｉｍａｘ可容许估计的充要条件

This paper discusses the Minimax admissibility characterization of linear estimates with respect to restricted regression coefficient under a matrix loss function. The necessary and sufficient conditions are given for a linear estimate LY to be Minimax admissibe in the class of homogeneous linear estimates.